Rahm Emanuel has a battle plan for Democrats, and it looks mighty familiar. On Tuesday, just hours before two special congressional elections were called for Republican candidates, the mayor of Chicago and political operative Bruce Reed published a piece in The Atlantic on “How the Democrats Can Take Back Congress.” As architects of the party’s midterm strategy in 2006, when Republicans lost control of both the House and Senate, they argued that “Donald Trump came to Washington to make waves—and he may deliver a wave election powerful enough to sweep his party out of control of Congress.” But, they added, “Waves don’t happen on their own: Democrats need a strategy, an argument, and a plan for what they’ll do if they win.”

Emanuel and Reed cautioned that the party can’t “rely entirely on one side’s enthusiasm or the other side’s disenchantment,” and that Democrats don’t need to spend the next year navel-gazing over how to motivate their base.” Instead, they need to “choose the right battles” and “choose credible candidates who can win them.” “Winning hotly contested swing seats,” they argued, “requires candidates who closely match their districts—even if they don’t perfectly align with the national party’s activist base.”

By definition, good candidates match their districts. But whether that requires deserting the “activist base” is an increasingly contested question in Democratic politics. In the wake of Bernie Sanders’s unexpected success in last year’s primary, and Hillary Clinton’s unexpected loss in the general election, the ascendant populist left is arguing that its policies are broadly popular—and that energizing the base is a more fruitful path to victory than fielding centrist candidates who can court Republicans or independents.

Assessing Democrat Jon Ossoff’s loss in Tuesday’s special congressional election in Georgia, some on the left blamed his centrist messaging—his focus on deficit reduction, for instance, and lack of anti-Trump rhetoric. “In the closing weeks of the race, Ossoff and the DCCC missed an opportunity to make Republicans’ attack on health care the key issue, and instead attempted to portray Ossoff as a centrist, focusing on cutting spending and coming out [in] opposition to Medicare for All,” Anna Galland, Move On’s executive director, said in a statement. “This approach did not prove a recipe for electoral success. Democrats will not win back power merely by serving as an alternative to Trump and Republicans.”

Ossoff, one could argue, was tailored to his district, and yet he fell short. It’s one of many reasons the Democratic establishment could face criticism if they pursue a midterm strategy similar to Emanuel’s, which drew objection from the left at the time—and still does.

Meredith Kelly, communications director at the Democratic Congressional Campaign Committee, says Ossoff didn’t have trouble courting progressives. “His base was there,” she told me on Wednesday. “I would argue there’s no evidence he had a problem on the progressive side.” Nor does she see evidence he should have been more overtly anti-Trump. “This is a district where Trump still is not underwater the way he is nationwide,” Kelly said. “This is a place where Trump was still 50-50 in terms of approval.” Ultimately, she told me, “There are going to be a lot of opinions. I don’t think anyone knows for sure why Jon Ossoff lost other than this district is really hard and he ran out of Democrats and independents to support him.”

Politico reported last month that national Democrats were consulting Emanuel on how to replicate his strategy for 2018: “Democrats believe President Donald Trump has already given them enough to make the ‘cronyism, corruption and incompetence’ argument they employed in 2006 — when [House Minority Leader Nancy] Pelosi and Senate Minority Leader Harry Reid first implored voters to ‘drain the swamp’ in Washington.” It’s hard to imagine any Democrat quibbling with that message—at least as part of their midterm argument—or objecting to the type of early and aggressive fundraising Emanuel employed in 2006. He relentlessly tied rank-and-file Republicans to President George W. Bush, and there’s no doubt Democrats should soil every GOP candidate with Trump’s disastrous agenda. (To be sure, Emanuel is also stressing that Democrats need their own positive agenda to compliment their negative messaging—another popular conclusion drawn from Tuesday’s results.)

Yet as much as 2006 was a historic victory for Democrats, Emanuel’s strategy prompted intra-party divisions—debates that remain relevant to the future of the party today. Wall Street Journal editor Naftali Bendavid covered the campaign for The Chicago Tribune and authored The Thumpin’: How Rahm Emanuel and the Democrats Learned to Be Ruthless and Ended the Republican Revolution. He told me, “There was also this huge debate going on, then as now, over whether you go with populist base-type candidates or recruit for the district,” which in 2006 meant running moderate and even conservative candidates. The debate then “was less economic populism, which seems to be part of the discussion now, but it had to do with gun control, abortion, and a certain degree of social conservatism,” Bendavid said. Victory in November didn’t resolve these arguments, either; some on the left argued they could have won bigger majorities with more progressive candidates.

In 2006, Democrats were also divided over Emanuel’s focus on swing districts versus Howard Dean’s push, as chair of the Democratic National Committee, for a “50-state strategy.” The disagreement resulted in a long-running feud between the two men, and both credited his own approach with the party’s midterm victory. Asked to comment on Emanuel’s Atlantic piece, Dean said, “I don’t think anyone outside the Beltway will read this or care.” He told me Emanuel made major contributions in 2006, particularly with fundraising, but Dean downplayed the extent to which Democrats can emulate that year’s strategy for 2018. “The playbook’s going to be very different,” he said. “To think that we’re going to use the strategy from 12 years ago that was important but not sufficient is silly.”

In Dean’s view, the DCCC and the DNC should leave candidate recruitment to people outside of Washington, support activist groups, and reengage millennials. “The activist base will support some moderate Democrats in the appropriate districts,” he stressed. “We have to trust these people.” If Democrats follow this approach, Dean is incredibly optimistic about their prospects. “I believe we’re going to take the Senate back in 2018 as well as the House, but we’re not going to do that if we start screwing around inside the Beltway, thinking we know best,” he said. “Washington does not understand what’s going on in the rest of the country, and if they did Donald Trump wouldn’t be president of the United States.”

Dean and others progressives also reject the idea that the DCCC’s strategy made the difference in 2006. Markos Moulitsas, founder and publisher of the progressive blog Daily Kos, told me, “We won in 2006 because George W. Bush had worn out out his welcome.” Moulitsas argues that voters were primed to oppose Republicans “no matter what warm body Democrats had thrown into a district,” and there’s a temptation to “way overplay the quality of a candidate in these wave scenarios.” Similarly, he said, “2018 will be a referendum on Trump, and the only thing that will matter for Democrats is having a ‘D’ next to their name. And given this year’s special election results, upward of 100 Republican-held seats could be in some level of play. We won’t win that many, for sure, but 40-60 isn’t out of the realm of possibility. And at that point, candidates aren’t winning based on their charming personalities and milquetoast politics. They are winning based on massive negative public sentiment.”

Moulitsas cautions Democrats not to depress the intensity of their base by softening their anti-Trump stance or trying to engineer candidates.If the base sees Senate Democrats slowing the Senate to a standstill to kill TrumpCare, they will be more motivated to work hard for Democrats next year. That’s what riding the wave looks like,” he said. “That means not overthinking how some candidate fits in with a district (often based on past voting patterns, not aspirational future ones based on non-voting potential base voters in a district), but rather, taking actions, starting today, that will further rev up base intensity, priming them for next year’s vital elections.”

“Midterm elections are overwhelmingly about turning out the base,” Ben Wikler, Move On’s Washington director, told me. Like Moulitsas, he thinks the “unified Democratic message about the culture of corruption” was effective a decade ago, and “gains that Democrats made in 2006 were the result of progressive organizing that capitalized on the public’s rejection of George W. Bush and cronyism.” “That is not a run-corporate-friendly-centrists strategy,” Wikler emphasized. “Our sense is that in 2018, even more than in 2006, the public is furious with politics run for the benefit of billionaires.... Especially in this moment of populist resistance energy, we’re going to be best served by candidates who unite progressives and the disaffected with authenticity and vision people can believe in.”

Kelly insists that the DCCC “has done a number of new things this cycle that get us outside the supposed bubble and closer to the ground on these districts,” including hiring local organizers to work with activists on the ground. She said “there are definitely parallels” with 2006, “but I think we’re also doing things in brand new fresh ways that are unique to the environment we’re in.” She added, “We’re in a new frontier and the grassroots are some of the most powerful people in our politics right now.”

“I think our candidates will be able to walk and chew gum at the same time in terms of appealing to Democratic base voters who are critical to winning in swing districts while also appealing to independents and some moderate Republicans,” Kelly said. “To Rahm’s point, these are Republican districts. Our whole battlefield is Republican-leaning.” She added that the party’s base “is not really imposing a litmus test as far as I’ve seen.”

The question is whether winning on a Republican-leaning battlefield requires centrists. “I’m not going to prescribe a specific ideology that a candidate needs,” Kelly said, “but what we do absolutely care about is that a candidate fits the district. We really want authentic people that understand these communities. There’s no one-size-fits-all-approach to these districts. We absolutely reserve the right to get involved in these primaries where necessary.” That’s certainly similar to the DCCC’s rationale in 2006. John Lapp, the committee’s executive director then, told me “2006 was less about an ideological profile for candidates and more about recruiting non-traditional candidates in non-traditional places. Helping recruit and win with sheriffs, military veterans, and even high school football coaches in purple and ruby red Republican districts all across the country. We’re certainly doing that now in 2017, as well.”

Everyone seems to agree on recruiting non-traditional candidates. “Donald Trump led the way in showing how the nation’s populist sentiment is anti-politician,” Moulitsas said. “Democrats would be wise to run veterans, teachers, firemen, mothers, and other such non-traditional faces. Let’s skip the typical politician, and let’s run people that voters can better identify with.” In a memo released after Ossoff’s loss, DCCC chairman Ben Ray Luján wrote, “Let’s look outside of the traditional mold to keep recruiting local leaders, veterans, business owners, women, job-creators, and health professionals. Let’s take the time to find people who fit their districts, have compelling stories, and work hard to earn support from voters.”

Where there’s likely to be debate, of course, is when these non-traditional candidates depart from progressive policies. In 2006, Bendavid said, “A lot of the people with military, police, and athletic backgrounds also were not down-the-line liberals.” Similarly, Luján’s memo argues that “the road back to a Democratic House majority ... necessitates fielding strong candidates with diverse profiles that fit unique Republican-leaning districts. It demands that we continue embracing a big tent mentality.”

The Democratic Party’s strategy aside, there’s already a sense that the grassroots are shaping the landscape more today than they did in 2006. Democrats are “busy sorting through potential candidates, who in some cases number more than a dozen interested prospects for a single district,” according to Politico. “The DCCC has been succeeding much earlier than usual in landing strong recruits.” “Democrats were largely demoralized after the 2004 election,” Lapp told me. “It took a real leap of faith for Democrats to believe again, particularly in these purple and red districts. We were desperate to have people run. That’s very different from 2017, where we’ve got an embarrassment of riches. Folks are coming out of the woodwork to run for Congress—all on their own.”

The Democratic Party will certainly play a major role in these races, as its various arms decide which candidates to allocate resources to and fundraise for. But it’s not clear that the DCCC, for instance, wields the same king-making power that it did under Emanuel. If the grassroots is indeed much stronger today than in 2006, they may well have the power to set the party on a new, leftward course, just as Sanders’s movement nearly did last year. As Lapp said, “It is up to Democratic primary voters to sort out who they prefer—pitchfork-wielding progressives or more moderate-minded candidates.”